The future intelligent transmitter will dynamically adjust the transmission configuration on demand, which bring new challenges to digital predistortion (DPD). In this article, we present a gated dynamic neural network (GDNN) DPD model linearize power amplifier (PA) with varying configurations. proposed GDNN is composed of gating and backbone that can be any NN-based designed for fixed configuration. core idea adjusted using configuration-dependent weights generated by achieve...

A novel behavioral modeling technique called pruned basis space search (PBSS) is proposed for digital predistortion (DPD) of RF power amplifiers (PAs). The PBSS finds the optimal DPD model by function in (PBS). PBS obtained sparsifying comprising a wide variety functions, while implemented based on heuristic algorithms. multiplexing-based complexity identification algorithm to improve fitness calculation so that can balance performance and running model. avoids shortcomings traditional...

Digital predistortion (DPD) has been widely used in linearizing radio frequency (RF) power amplifiers (PAs). However, model coefficients could not always be estimated accurately for a variety of reasons. Several regularization methods have developed parameter identification. the performance improvement is limited due to missing information. Fortunately, if parameters from earlier operating conditions are available, they can employed enhance accuracy DPD current state. Despite fact that many...

Digital predistortion (DPD) is an effective linearization technique for RF power amplifiers (PAs), but conventional full sampling (FS) DPD systems use ADCs with three to five times signal bandwidth, and high-speed are expensive power-hungry. In this article, we develop a novel band-limited reducing feedback rate acquisition bandwidth based on general framework semisupervised learning called manifold regularization (MR), which utilizes the geometry of unlabeled data construct terms mitigating...

In this letter, two basis function multiplexing-based behavioral modeling methods for digital predistortion (DPD) of RF power amplifiers (PAs) are proposed to reduce the running complexity DPD. The full basis-propagating selection (FBPS) model and reduced-complexity FBPS (RC-FBPS) give reasonable ways multiplex even-order functions, extending (BAPS) which only uses delay odd-order functions. experimental results confirm that both RC-FBPS models can achieve a good tradeoff between performance.

In this letter, we propose a method for behavioral modeling and digital predistortion (DPD) of RF power amplifiers (PAs) based on multi-output recurrent neural networks (RNNs). RNN has high accuracy, but it also running complexity due to the mechanism. For reason, model architecture, which means that DPD produces multiple adjacent outputs simultaneously single input sample group. This approach greatly reduces RNNs with essentially no deterioration in performance. The proposed mechanism is...

To accomplish rapid adaptation of the digital predistortion (DPD) model, a low-complexity parameter extraction architecture is proposed in this article. The extracted DPD model coefficients are represented by linear combination previous parameters (or pretrained parameters) novel basis (BPC) method, thereby avoiding high-dimensionality and significantly lowering computational cost. Then, we developed feature mapping technique (FMT) to coordinate spaces corresponding different structures,...

A novel behavioral modeling approach called adaptive model tree (AMT) is proposed for digital predistortion (DPD) of RF power amplifiers (PAs) in fixed and time-varying configurations. The AMT piecewise based on the decision reduced-complexity full basis-propagating selection (RC-FBPS) model. two-step joint iterative algorithm to achieve a good match between submodels obtained from RC-FBPS inherits enhances respective advantages have powerful capability potentially. experimental tests...

This article proposed a novel sample selection strategy for reducing the computational complexity of digital predistortion (DPD). Due to memory effect power amplifier (PA), PA's output is affected by term. Thus, unlike existing methods (SSMs) that consider signal amplitude as only feature, method regards points and their lagged terms (memory terms) features each point. We also introduce representative subset further increase selected samples' diversity, these are improved reduce storage...

In this paper, we present a novel behavioral modeling technique based on decomposed vector combination (DVC) for digital predistortion (DPD) of RF Transmitters. The basis function the proposed DVC model consists piecewise function-based magnitude term and linear phase-combination-based phase term. functions are still linear-in-parameters go beyond classical Volterra series. Compared with DPD models, has theoretically more powerful capability through richer form functions. A...

In this article, a novel block-oriented recurrent neural network (RNN) model is proposed for behavioral modeling and digital predistortion (DPD) of radio frequency (RF) power amplifiers (PAs). This article provides an insightful discussion on the importance input-end parallel finite impulse response (FIR) filters performance enhancement finds, first time, unique linearization correction effect each FIR filter in at different frequencies, which also reason why time-delay NN (BOTDNN)...

A simplified adaptive model tree (SAMT) for over-the-air (OTA) behavioral modeling of millimeter-wave (mmWave) beamforming transmitters in time-varying transmission configurations is proposed this paper. The SAMT significantly reduces its algorithm complexity and running by simplifying sub-models the AMT to overcome problem high when applied mmWave transmitters. inherits advantages can behavior main beam transmitter different directions using only one model. experimental results confirm that...

In this article, we present a low-complexity adaptive digital predistortion (DPD) model for concurrent multiband (MB) power amplifiers (PAs). A novel MB basis function space is constructed the proposed vector combination search (MB-VCS) by performing multidimensional extensions and adding piecewise linear functions to Polar Volterra series. The multiplexing-based algorithm also helps MB-VCS balance performance complexity. contains in-band cross-band terms characterize complex interactions...

In this paper, a neural-network (NN) based digital predistortion (DPD) model is proposed for the fully-connected (FC) hybrid beamforming (HBF) massive MIMO (mMIMO) transmitters. Comparing to existing models, can effectively linearize virtual main far-field signal and achieves better performance. Simulations on 2-stream 64-element FC HBF array demonstrate effectiveness of DPD models.

In this paper, a new 1-bit vector-switched (VS) model with localized step size is proposed to improve the linearization performance of existing digital predistortion (DPD) model. The method considers iteration calculation errors in different partitions input signal amplitude space. experiment results show that has better than DPD and more comparable conventional high-precision model, especially when power amplifier exhibits unusual nonlinearity.